openskill

library6.2.0pypypi✓ verified 85d ago

Multiplayer rating system implementing the Weng-Lin models (Plackett-Luce, Bradley-Terry, Thurstone-Mosteller). Pure Python, no dependencies. Version 6.2.0 requires Python ~=3.10. Active development with regular releases.

ai-ml

pip install openskill

github ↗package ↗docs ↗changelog ↗

INSTALL

IMPORT

SIG · OPENSKILL

openskill

ai-mlpythonv6.2.0

Install

1.7s avg

Import

10ms

Disk

18MB

Pass rate

10/ 10

Env Coverage10 / 10

glibc

3.9–3.13

musl

3.9–3.13

Install & Compatibility

Where this runs

tested against v6.2.0 · pip install

no network on importno background threads

Install × environment matrix

Each cell = how many times install + import succeeded across repeated harness runs. Partial = flaky.
glibc = Debian/Ubuntu slim · musl = Alpine Linux

musl

py 3.10–3.95 runs

installs and imports cleanly · install 0.0s · import 0.002s · 20.3MB

glibc

py 3.10–3.95 runs

installs and imports cleanly · install 1.7s · import 0.000s · 21MB

18MB installed

● package 18MB

Code

Verified usage

Verified import paths — ran on the pinned version, not inferred.

openskill

✓ import openskill

✗ from openskill import *

Wildcard imports are discouraged; import the module and use its functions.

rate

✓ openskill.rate(t1, t2)

Simple example: two players, one match, team1 wins.

import openskill # Define two teams with one player each team1 = [openskill.Rating(mu=25, sigma=8.33)] team2 = [openskill.Rating(mu=25, sigma=8.33)] # Team1 wins (lower rank is better, ranks default to 0,1,...) team1, team2 = openskill.rate([team1, team2], ranks=[0, 1]) print(team1[0]) print(team2[0])

Debug

Known issues

breakingIn v6.0.0, the `limit_sigma` parameter is no longer implicitly set; you must pass it explicitly if you want sigma to have a floor.

fix

Add `limit_sigma=True` to your rating or create Rating objects with `limit_sigma` parameter.

affects: >=6.0.0

breakingv6.1.3 fixed a critical bug where players could gain infinite rating under certain conditions. Ensure you are on at least 6.1.3.

fix

Upgrade to openskill>=6.1.3.

affects: <6.1.3

deprecatedPython 3.9 support dropped in v6.0.1. Use Python 3.10+.

fix

Use Python >=3.10.

affects: >=6.0.1

gotchaThe `rate` function expects a list of lists (teams) and returns updated teams. Mutating the input objects may lead to unexpected results.

fix

Treat tuples as immutable; do not modify original Rating objects after passing to rate.

affects: all

gotchaRanks are zero-indexed (lower rank = better). If you pass ranks incorrectly, win/loss direction may invert.

fix

Use ranks=[0,1] where team1 is better (lower rank) than team2.

affects: all

Errors

Common errors & fixes

ImportError: cannot import name 'rate' from 'openskill'

Trying to import rate as a top-level function directly, but it's a module function.

fix

Use `import openskill` then `openskill.rate(...)`.

AttributeError: module 'openskill' has no attribute 'Rating'

typo: 'Rating' not 'rating'.

fix

Use `openskill.Rating()` (capital R).

ValueError: A team should have at least one player

Passing an empty list as a team.

fix

Ensure each team has at least one Rating object.

TypeError: __init__() got an unexpected keyword argument 'limit_sigma'

Using a pre-6.0.0 version where limit_sigma wasn't a parameter.

fix

Upgrade to openskill>=6.0.0 or omit the parameter.

Upgrade

Version history

6.2.0latest on PyPI · released Jan 11, 2026

Audit

Dependencies

No dependency data recorded yet.

Agent activity

5 hits · last 30 days

node

Amazon

Resources

githubhttps://github.com/vivekjoshy/openskill.py ↗

packagehttps://pypi.org/project/openskill/ ↗

docshttps://openskill.me/en/stable ↗

changeloghttps://github.com/vivekjoshy/openskill.py/blob/main/CHANGELOG.rst ↗